Transcript The Stromgren sphere around the highest-redshift QSOs

Evolution of Accretion Disks around
Massive Black Holes: Constraints from
the Demography of Active Galactic Nuclei
Qingjuan Yu
UC Berkeley
April 21, 2006
(2005, ApJ, 634, 901, Qingjuan Yu, Youjun Lu, & Guinevere Kauffmann)
Introduction
• QSOs are powered by gas accretion onto MBHs.
• Most nearby galaxies host MBHs at their centers.
• Mass growth of MBHs comes mainly from gas
accretion due to QSO/AGN phases.
(Lynden-Bell 1969; Rees 1984; Soltan 1982; Small & Blandford 1992;
Kormendy & Richstone 1995; Magorrian et al. 1998; Yu & Tremaine
2002 etc.)
(Tremaine et al. 2002)
NGC 4258
Quasar PKS 2349 (HST)
Quasar PKS 2349
M87 (HST)
(HST)
Galactic center
M87
(HST)
• How does the accretion/luminosity evolve?
()
Quic kTime™ and a
TIFF (Unc ompres sed) dec ompres sor
are needed to see this pic ture.

Evolution after the nuclear activity
of a QSO/AGN is triggered
(1   )L( )
M&( ) 
 c2

M ( )  M  M&( ')d '
i

0
Not meaning
• Evolution of the
characteristic luminosity
of the QSO population:
• Cosmological evolution
of comoving number
density of the QSO
population:
Extracting evolution of accretion from observations
Statistical methods involving a large
sample of QSOs/AGNs are required.
2dF
SDSS
• A single AGN may only
represent one specific period in
a prolonged phase of nuclear
activity.
• A large sample of AGNs with
different ages will span all
phases of this activity and allow
us to extract information about
evolution.
• In addition to age, other
physical parameters may be
important in determining how
AGNs evolve, and a statistical
method may help to clarify
these.
Extracting
()
• Local BHs with present-day mass M0:
– Triggering history: seed BHs triggered at
cosmic time ti;
– Luminosity evolution (M0,) as a
function of =t-ti;
•

O
QSOLF
t
(M0,) is isolated by connecting QSOLF with local BHs:
(ignoring BH mergers)

t0
0

 (L,t)dt   nM ( M 0 ,t0 ) life ( M 0 )P(L | M 0 )dM 0
0
QSOLF
local BHMF lifetime probability
(Yu & Lu 2004)
Luminosity evolution of individual triggered nuclei
(M0,)
L+dL
L
seed BH
triggered
life


t0
0

 (L,t)dt   nM ( M 0 ,t0 ) life ( M 0 )P(L | M 0 )dM 0
0
QSOLF
local BHMF
lifetime probability
P(L | M 0 ) or
(M0,)
 life ( M 0 )P(L | M 0 )
L+dL
L
seed BH
triggered


t0
0

 (L,t)dt   nM ( M 0 ,t0 ) life ( M 0 )P(L | M 0 )dM 0
0
QSOLF
local BHMF
lifetime probability
Accretion rate distribution of SDSS nearby AGNs
(Yu, Lu & Kauffmann 2005)
Accretion rate distribution of SDSS nearby AGNs
Normalized mass accretion rate:
L[OIII]
m&[OIII]  f
LEdd (M f )
f : average bolometric correction
between L[OIII] and Lbol ;
M f ( ) : average final mass.
SDSS sample: (Kauffmann et al. 2003; Heckman et al. 2004)
z  0.3;
binning m[OIII] and  ;
 : 70  200km / s; M f ( ) : 2.0  10 6  1.3  10 8 M sun .
Accretion rate distribution of SDSS nearby AGNs
Accretion rate evolution M&bol ( )
P( M&bol | M f )d log10 M&bol
( M&bol ln10)d log10 M&bol

k
dM& ( ) d
bol
  k
 k : solutions of M&bol ( )  M&bol
(k  1,2,...).
-Assumed accretion rate evolution:

  
0    I;
exp   ,

 Sp 

M&  

    I   D 
,
  I.


D


I
II
I

Accretion rate distribution of SDSS nearby AGNs
-Assumed accretion rate evolution:

  
0    I;
exp   ,

 Sp 

M&bol  

    I   D 
,
  I.


D


I
I
  1.3  0.1,
II   3.1  1 .
D
Sp

(Yu, Lu & Kauffmann 2005)
Evolution model of accretion disks:
• Evolution of surface mass density:
 3   1/2 
1/2 

R
(

R
) ,
 R R 
R

   m Rn ;
• Self-similar solutions (Pringle 1974):


(R,  )     R     
   f     
0
  0   R0    0  
M&disk  

38 18a  4b
32 17a  2b
 1.18 , (a  b  0; Thomson opacity);
  1.25
, (a  1,b  7 / 2;Kramers opac.)

opacity :  ( ,T )   aT b
(Cannizzo, Lee, & Goodman 1990)
Evolution model of accretion disks:
• Diffusion timescale


R02
R0
0 
 (0.1  1.6)  10 8 yr 
 0.3  1pc 
 (R0 ,  0 )
7/3
 M BH 
 10 7 M 
sun
1/ 3
  


0.1 
4 / 3
 M d,0 
 10 7 M 
sun
2 / 3
• Consistency of observations with simple theoretical
expectations suggests that the accretion process in nearby
AGNs follows a self-similar evolutionary pattern.
T Tauri star
M&disk   
• Disk accretion: self-similar
evolution
(Hartmann et al. 1998)
Diversity of Eddington ratios (Lbol/Ledd) in QSOs/AGNs
QuickTime™ and a
TIFF (Uncompressed) decompre
are needed to see this pictur
(Mclure & Dunlop 2004)
The diversity in the Eddington ratios is
a natural result of the long-term evolution
of accretion disks in AGNs.
(Woo & Urry 2002)
Discussions
• Further issues related to long-term evolution of accretion
disks:
– Disk winds, infalling material deposited onto the
disk, instabilities, self-gravitating disks, star
formation …
• Binary black holes and coevolution of galaxies and
QSOs/AGNs
Discussions
• Adding the effect of an
evolving accretion disk in
unified models of AGNs
– Lack of a torus in very
weak AGNs
– Radiatively inefficient
accretion
Summary
• The accretion rates in most nearby Seyfert galaxies (with
host galaxy velocity dispersion sigma~70-200km/s, z<0.3)
are declining with time in a power-law form and the
accretion process follows a self-similar evolutionary
pattern as simple theoretical models predict.
• Some other issues deserves of further investigation, such
as the long-term evolution of accretion disks, the evolution
of BBHs in QSOs/AGNs, coevolution of galaxies and
QSOs/AGNs, and the unification picture of AGNs.
Alternative explanation for the accretion
rate distribution
• Fueling low-level AGN activity through the
stochastic accretion of cold gas,
astro-ph/0603180, Hopkins & Hernquist
– Feed-back driven model in a large-scale context
But how can the evolution of accretion disks
be avoidable?